US11423099B2 - Classification apparatus, classification method, and classification program - Google Patents
Classification apparatus, classification method, and classification program Download PDFInfo
- Publication number
- US11423099B2 US11423099B2 US16/955,083 US201816955083A US11423099B2 US 11423099 B2 US11423099 B2 US 11423099B2 US 201816955083 A US201816955083 A US 201816955083A US 11423099 B2 US11423099 B2 US 11423099B2
- Authority
- US
- United States
- Prior art keywords
- websites
- abused
- result data
- access result
- website
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active, expires
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/50—Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems
- G06F21/55—Detecting local intrusion or implementing counter-measures
- G06F21/556—Detecting local intrusion or implementing counter-measures involving covert channels, i.e. data leakage between processes
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/951—Indexing; Web crawling techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/958—Organisation or management of web site content, e.g. publishing, maintaining pages or automatic linking
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F17/00—Digital computing or data processing equipment or methods, specially adapted for specific functions
- G06F17/10—Complex mathematical operations
- G06F17/18—Complex mathematical operations for evaluating statistical data, e.g. average values, frequency distributions, probability functions, regression analysis
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F21/00—Security arrangements for protecting computers, components thereof, programs or data against unauthorised activity
- G06F21/50—Monitoring users, programs or devices to maintain the integrity of platforms, e.g. of processors, firmware or operating systems
- G06F21/55—Detecting local intrusion or implementing counter-measures
- G06F21/56—Computer malware detection or handling, e.g. anti-virus arrangements
- G06F21/562—Static detection
- G06F21/563—Static detection by source code analysis
-
- G06K9/6267—
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L67/00—Network arrangements or protocols for supporting network services or applications
- H04L67/01—Protocols
- H04L67/02—Protocols based on web technology, e.g. hypertext transfer protocol [HTTP]
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1408—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic by monitoring network traffic
- H04L63/1416—Event detection, e.g. attack signature detection
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1433—Vulnerability analysis
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L63/00—Network architectures or network communication protocols for network security
- H04L63/14—Network architectures or network communication protocols for network security for detecting or protecting against malicious traffic
- H04L63/1441—Countermeasures against malicious traffic
- H04L63/1483—Countermeasures against malicious traffic service impersonation, e.g. phishing, pharming or web spoofing
Definitions
- the present invention relates to a classification apparatus, a classification method, and a classification program.
- Website defacement where websites of others are abused has been a problem.
- An attacker achieves his/her goal by, for example, defacing another person's website so that transfer to a malicious website is carried out, the malicious website being, for example, a website that launches a drive-by download attack or a phishing site.
- Such website defacement is caused by: abuse of vulnerability existing in a web server or a web application; or theft of website management information through malware infection.
- CMSs content management systems
- a method of finding a website that is managed by a CMS of an old version by using the Dork technique and alerting an administrator of the website is one of simple methods of carrying out such alerting.
- the administrator may be managing and operating the website appropriately and using the old version of the CMS intentionally for some reason. Therefore, just because a CMS used for a website is of an old version, the website cannot be said to have, as a rule, a possibility of being abused.
- Non-Patent Literature 1 Methods of detecting drive-by download attacks, phishing, and hacktivism on websites have been known conventionally (as seen in Non-Patent Literature 1 to Non-Patent Literature 3). Furthermore, a method of designing feature values from a website before abuse and detecting the abuse of the website based on the feature values beforehand has been known (as seen in Non-Patent Literature 4). In all of the above mentioned methods, URLs and web content of websites are mainly used in designing of feature values useful for detection of abuse of a website.
- an object of the present invention is to solve the above described problem to accurately detect beforehand a website having a possibility of being abused due to a state of management of the website by an administrator thereof.
- the present invention is a classification apparatus, comprising: an acquisition unit that acquires access result data for websites updated by administrators of the websites, the access result data including information on the websites; an update history generation unit that generates update histories of the websites by making comparison between sets of access result data for the websites in chronological order; and a classifier generation unit that generates a classifier that classifies a website having a possibility of being abused in the future by using feature values that are the update histories of the websites.
- the present invention enables accurate detection of a website beforehand, the website having a possibility of being abused due to a state of management of the website by an administrator thereof.
- FIG. 1 is a diagram illustrating an outline of operation of a classification apparatus.
- FIG. 2 is a block diagram illustrating an example of a configuration of the classification apparatus.
- FIG. 3 is a diagram illustrating an example of an access result DB in FIG. 2 .
- FIG. 4 is a diagram for explanation of generation of update histories.
- FIG. 5 is a diagram illustrating an example of comparison between sets of access result data.
- FIG. 6 is a diagram illustrating examples of update histories.
- FIG. 7 is a diagram illustrating an example of a feature vector representing the update history in FIG. 6 .
- FIG. 8 is a flow chart illustrating an example of a procedure for generation of a classifier by the classification apparatus in FIG. 2 .
- FIG. 9 is a flow chart illustrating an example of a procedure for classification by the classification apparatus in FIG. 2 .
- FIG. 10 is a flow chart illustrating an example of processing for accumulation of access result data into the access result DB by the classification apparatus in FIG. 2 .
- FIG. 11 is a flow chart illustrating an example of processing for generation of an update history by the classification apparatus in FIG. 2 .
- FIG. 12 is a diagram illustrating a computer that executes a classification program.
- a classification apparatus constantly monitors websites and collects information considered to be attributable to administrators thereof from the websites. The classification apparatus then classifies websites that may be abused in the future by chronologically analyzing how the monitored websites have been managed and operated retroactively.
- the classification apparatus accesses websites included in URL lists for learning (an unmalicious URL list for learning and a malicious URL list for learning) via the Internet through a crawler unit (acquisition unit) 121 .
- the crawler unit 121 collects access result data for the websites and accumulates the access result data into an access result DB.
- the access result data for the websites include information updated due to management by administrators of the websites (for example, web application information described later).
- URL lists provided by service for archiving past states of various websites are used as the unmalicious URL list for learning and the malicious URL list for learning.
- a URL list provided by service for archiving states of websites at different times after detection of abuse and before the detection of abuse is used as the malicious URL list for learning.
- a URL list provided by service for archiving states of websites at different times, the websites not having been abused, is used as the unmalicious URL list for learning.
- the classification apparatus generates update histories of the websites by making comparison between sets of access result data for each website, the sets of access result data having been accumulated in the access result DB.
- the classification apparatus generates an update history of a website that has been abused, based on access result data before detection of the abuse from access result data for the websites included in the malicious URL list for learning. That is, the update history of the website prior to the detection of the abuse is generated. Furthermore, through the update history generation unit 122 , the classification apparatus generates an update history of a website that has not been abused, based on access result data for the websites included in the unmalicious URL list for learning.
- a classifier generation unit 123 of the classification apparatus then generates a classifier that classifies websites having a possibility of being abused in the future by performing machine learning or the like through use of feature values that are the update histories of the websites that have been abused and the update histories of the websites that have not been abused.
- the classification apparatus performs classification of websites included in a URL list for testing (a list of URLs of websites to be classified) by using the classifier that has been generated.
- the classification apparatus accesses, through the crawler unit 121 , the websites included in the URL list for testing, and accumulates access result data for the websites into the access result DB.
- the classification apparatus makes comparison between sets of access result data for each website, the sets of access result data having been accumulated in the access result DB, and generates update histories of the websites. Thereafter, the classification apparatus performs classification of whether or not each website included in the URL list for testing has a possibility of being abused in the future, by using the classifier and feature values that are the update histories of the websites included in the URL list for testing, and outputs a result of the classification.
- the classification apparatus is able to accurately detect beforehand a website having a possibility of being abused due to a state of management of the website by an administrator thereof.
- the classification apparatus 10 includes an input and output unit 11 , a control unit 12 , and a storage unit 13 .
- the input and output unit 11 is an input and output interface that receives input of various URL lists and outputs, for example, a result of classification by a classifier, for websites represented by URLs. Furthermore, this input and output unit 11 serves as a communication interface for the crawler unit 121 or the like to make access to an external network, such as the Internet.
- the control unit 12 controls the overall classification apparatus 10 .
- This control unit 12 includes the crawler unit 121 , the update history generation unit 122 , the classifier generation unit 123 , and a classification unit 124 .
- the crawler unit 121 makes access to websites of specified URLs and collects (or acquires) various types of information. For example, at predetermined intervals, the crawler unit 121 makes access to websites included in a URL list, collects access result data for the websites accessed, and accumulates the access result data into the access result DB.
- the access result data include information on the websites (for example, the web application information described later) updated due to management by administrators of the websites.
- the crawler unit 121 is implemented by, for example, a browser for accessing the websites, or execution of a crawler program.
- URL lists input to the crawler unit 121 are, for example, the URL lists for learning (the unmalicious URL list for learning and the malicious URL list for learning) and the URL list for testing.
- URL lists provided by service (Document 1) for archiving past states of websites not detected to have been abused (unmalicious websites) and websites detected to have been abused (malicious websites) as described above, for example, are used as the unmalicious URL list for learning and the malicious URL list for learning.
- a malicious URL list including URLs of websites detected beforehand to have been abused may be used, for example, by use of an existing technique (for example, anti-virus software or a technique described in Non-Patent Literature 1, 2, or 3).
- the crawler unit 121 makes access to websites that have been archived at times before detection of abuse and collects access result data.
- the crawler unit 121 then accumulates the access result data collected, as access result data before detection of abuse (a series of access result data prior to the detection of abuse), into the access result DB.
- the crawler unit 121 may constantly access a website represented by a specific URL for a predetermined period, and if the crawler unit 121 detects abuse of the website through the existing technique, the crawler unit 121 may use access result data that have been collected prior to that detection as access result data before detection of abuse.
- the access result data for the website collected by the crawler unit 121 include information composing the website.
- the access result data for the website include at least one of: web server information on the website acquired from an HTTP header of the website; web application information on the website acquired from an HTTP body of the website (for example, a meta tag of the HTTP body); and CMS information on the website.
- the access result data include web server information acquired from a server header or an X-Powered-By header included in an HTTP header, web application information acquired from a meta tag included in an HTTP body, and CMS information acquired from a path string of the URL.
- the CMS information is, for example, information indicating whether or not a CMS of the website is able to be identified from each of pieces of information that are the URL of the website, the HTTP header, and the meta tag of the HTTP body.
- CMS information is information, such as “detect-url” indicating identifiability of a CMS from a URL, “detect-meta” indicating identifiability of the CMS from a meta tag, and “detect-header” indicating identifiability of the CMS from an HTTP header.
- web server information is information, such as “Server header” and “X-Powered-By header”.
- web application information is information, such as “meta[generator]” that is a generator attribute value of the meta tag.
- the classification apparatus 10 uses, as a feature value, information that is able to be updated by an administrator of a website, such as web server information on the website or web application information, as described above.
- the feature value may be any information that is able to be updated by the administrator of the website, and is not necessarily the web server information or web application information described above.
- Access result data accumulated in the access result DB will be described by use of FIG. 3 .
- the access result data accumulated in the access result DB include information, such as: “ID” that is identification information of each set of access result data; “Date” that is a date and time of access to a website; “URL” that is a URL of the website that has been accessed; “Number of URLs” that is the number of URLs that are additionally accessed when the URL is accessed; “meta[generator]” that is a generator attribute value of a meta tag of an HTTP body; “Server header” and “X-Powered-By header” included in an HTTP header; “detect-url” indicating whether or not identification of a CMS is possible from the URL; “detect-meta meta” indicating whether or not identification of the CMS is possible from the meta tag; and “detect-header” indicating whether or not identification of the CMS is possible from the HTTP header.
- ID identification information of each set of access result data
- values of “detect-url”, “detect-meta”, and “detect-header”, that is, “True” (identification of the CMS being possible) or “False” (identification of the CMS not being possible), in the above described access result data, are determined by the crawler unit 121 , based on an existing technique (for example, a technique described in Document 4 cited below).
- the crawler unit 121 When the crawler unit 121 makes access to a website included in a URL list for learning, the crawler unit 121 : assigns identification information to access result data for the website, the identification information indicating whether the website to be accessed is of an unmalicious URL (a website that has not been abused) or of a malicious URL (a website that has been abused); and accumulates the access result data assigned with the identification information, into the access result DB.
- an unmalicious URL a website that has not been abused
- a malicious URL a website that has been abused
- the update history generation unit 122 in FIG. 2 Based on the access result data accumulated in the access result DB (as seen in FIG. 3 ), the update history generation unit 122 in FIG. 2 generates an update history of a website. For example, the update history generation unit 122 acquires, from the access result DB, sets of access result data for the same URL (called a seed URL) and makes comparison between the sets of access result data in chronological order. For example, the update history generation unit 122 makes comparison between pieces of information included in the sets of access result data to be compared (for example, comparison between character strings and numerical values, and comparison between truth values), and stores results of the comparison as an update history of the website of the seed URL, into the storage unit 13 .
- a seed URL sets of access result data for the same URL
- the update history generation unit 122 makes comparison between pieces of information included in the sets of access result data to be compared (for example, comparison between character strings and numerical values, and comparison between truth values), and stores results of the comparison as an update history of the website of the seed URL
- the update history generation unit 122 When generating an update history of a website included in the malicious URL list for learning, the update history generation unit 122 generates the update history of the website that has been abused, based on access result data before detection of abuse of the website from access result data for the website.
- the update history generation unit 122 uses the access result data before detection of abuse (access result data not detected to have been abused) from the access result data for the website for the dates and times, in generating an update history of the website that has been abused. That is, for the website that has been abused, the update history generation unit 122 generates the update history of the website by using the access result data for dates and times prior to the detection of abuse.
- the update history generation unit 122 when generating an update history of a website included in the unmalicious URL list for learning (a website that has not been abused), the update history generation unit 122 generates the update history of the website by using all of access result data for the website (access result data for every date and time).
- the generation of the update history by the update history generation unit 122 will be described by use of a specific example.
- the update history generation unit 122 records “Up” if the character string has become observable, “Stable” if the character string has not changed, “Updated” if the character string has changed, and “Down” if the character string has become unobservable, into an update history.
- the update history generation unit 122 records “Up” if the numerical value has increased, “Stable” if the numerical value has not changed, and “Down” if the numerical value has decreased, into the update history.
- the update history generation unit 122 records “Up” if the truth value has changed from falsity to truth, “Stable” if the truth value has not changed, and “Down” if the truth value has changed from truth to falsity, into the update history.
- the update history generation unit 122 may use numerical values or character strings other than “Up”, “Stable”, “Updated”, and “Down”, as long as a change in value is able to be represented uniquely.
- the update history generation unit 122 accepts data types other than those of character strings, numerical values, and truth values, according to data collected by the crawler unit 121 .
- the update history generation unit 122 may record a time interval (the number of months) between access dates and times for the seed URL into the update history (as seen in FIG. 6 as “Delta”), the time interval having been acquired by comparison between values for “Date” in the access result data as illustrated in FIG. 3 .
- the classifier generation unit 123 is able to generate a classifier in consideration of a feature value that is the time interval between the access dates and times for the seed URL when the classifier generation unit 123 generates the classifier.
- Classification accuracy for websites that may be abused in the future is able to be improved by the classification unit 124 receiving input of an update history including a time interval between access dates and times for a website to be classified and performing classification by using a classifier as described above.
- the update history generation unit 122 may generate an update history by using an older set of access result data for the same seed URL.
- Update histories used by the classifier generation unit 123 in generating a classifier each need to be an update history that has been acquired by comparison between sets of access result data for the same seed URL for the same number of times. Therefore, the update history generation unit 122 executes the following processing in order to make the number of times comparison is made between sets of access result data the same for all of the update histories of the access result data used in generating a classifier.
- an update history used in generation of a classifier needs to be an update history acquired by comparison between sets of access result data for three times, but only three sets of access result data for the same seed URL have been accumulated in the access result DB (that is, comparison is able to be made only twice).
- the update history generation unit 122 records, as a result of comparison for the third time, a missing value (for example, “ ⁇ 1”) into the update history.
- ⁇ 1 is used as the missing value in the above described example, any value other than “ ⁇ 1” may be used as long as the value is able to be determined as the missing value.
- the update history generation unit 122 is able to generate update histories resulting from comparison between sets of access result data for the same number of times for each of the update histories used in generation of a classifier. Any value may be set as the number of times comparison between sets of access result data is made.
- the classifier generation unit 123 in FIG. 2 generates a classifier that classifies websites having a possibility of being abused in the future, by performing machine learning or the like using feature values that are update histories of websites that have not been abused and update histories of websites that have been abused.
- the classifier generation unit 123 reads out update histories of websites included in the unmalicious URL list for learning and update histories of websites included in the malicious URL list for learning, both of which are stored in the storage unit 13 .
- the classifier generation unit 123 then generates feature vectors representing the update histories of the websites included in the unmalicious URL list for learning and feature vectors representing the update histories of the websites included in the malicious URL list for learning.
- the classifier generation unit 123 generates a classifier that classifies websites having a possibility of being abused in the future, by performing machine learning or the like using unmalicious data that are the feature vectors representing the update histories of the websites included in the unmalicious URL list for learning and malicious data that are the feature vectors representing the update histories of the websites included in the malicious URL list for learning.
- the classifier generation unit 123 generates a feature vector illustrated in FIG. 7 as a feature vector representing the update history for the seed URL “http://c.example” illustrated in FIG. 6 (the case where the number of times comparison is made is three). Although illustration thereof is omitted, the feature vector has been assigned with identification information indicating whether the feature vector is a feature vector of malicious data or a feature vector of unmalicious data. The classifier generation unit 123 then generates a classifier that classifies websites having a possibility of being abused in the future by performing machine learning or the like by using feature vectors including the identification information.
- the classification unit 124 in FIG. 2 classifies websites having a possibility of being abused in the future by using the classifier generated by the classifier generation unit 123 , and outputs a result of the classification. For example, the classification unit 124 performs classification of whether or not each website included in the URL list for testing has a possibility of being abused in the future by using feature vectors representing update histories of websites included in the URL list for testing and the classifier generated by the classifier generation unit 123 , and outputs a result of the classification.
- FIG. 8 Next, an example of a procedure of processing by the classification apparatus 10 will be described by use of FIG. 8 and FIG. 9 .
- a procedure where the classification apparatus 10 generates a classifier will be described by use of FIG. 8 .
- the crawler unit 121 of the classification apparatus 10 makes access to websites included in the unmalicious URL list for learning and malicious URL list for learning (S 1 ), and accumulates access result data acquired, into the access result DB (S 2 ). Thereafter, the update history generation unit 122 generates an update history by acquiring, in chronological order, sets of access result data related to the same seed URL that have been accumulated in the access result DB (S 3 ). The classifier generation unit 123 then generates a classifier that performs classification of whether or not each website has a possibility of being abused in the future by performing machine learning or the like using the update histories related to the websites included in the unmalicious URL list for learning and the update histories related to the websites included in the malicious URL list for learning (S 4 ).
- FIG. 9 Described next by use of FIG. 9 is a procedure in which websites having a possibility of being abused in the future are classified, for websites included in the URL list for testing, by use of the classifier generated by the classification apparatus 10 .
- the crawler unit 121 of the classification apparatus 10 makes access to a website included in the URL list for testing (S 11 ), and accumulates access result data acquired through the access, into the access result DB (S 12 ).
- the update history generation unit 122 generates an update history by acquiring, in chronological order, sets of access result data related to the same seed URL, the sets of access result data having been accumulated in the access result DB (S 13 ).
- the classification unit 124 performs classification of URLs for testing, the URLs being included in the URL list for testing (S 14 ).
- the classification unit 124 performs classification of whether or not the website included in the URL list for testing is a website having a possibility of being abused in the future.
- the classification unit 124 then outputs a result of the classification (a classification result) of S 14 (S 15 ).
- the classification apparatus 10 is able to accurately detect beforehand a website having a possibility of being abused due to a state of management of the website by an administrator thereof.
- the crawler unit 121 acquires a URL to be accessed next from a URL list that has been input (S 21 ), makes access to the acquired URL to be accessed, and registers access result data that have been acquired, into the access result DB (S 22 ). If the crawler unit 121 then determines that the crawler unit 121 has made access to all of URLs included in the URL list (Yes at S 23 ), the crawler unit 121 ends the processing, and if there is a URL that has not been accessed yet in the URL list (No at S 23 ), the crawler unit 121 returns to S 21 .
- the update history generation unit 122 attempts to acquire, in chronological order, access result data related to a seed URL from the access result DB (S 31 ), and if the access result data have been able to be acquired (Yes at S 32 , “Are access result data [acquired] present?”), the update history generation unit 122 proceeds to S 33 . On the contrary, if access result data related to the seed URL have been unable to be acquired, in chronological order, from the access result DB (No at S 32 , “Are access result data [acquired] present?”), the update history generation unit 122 records a missing value (for example, “ ⁇ 1”) into the update history (S 34 ) and proceeds to S 38 .
- a missing value for example, “ ⁇ 1”
- the update history generation unit 122 checks for presence of access result data (access result data [held]) that have been used in the last comparison of access result data, and if the access result data [held] are present (Yes at S 33 ), the update history generation unit 122 proceeds to S 35 . On the contrary, if the access result data [held] are not present (No at S 33 ), the update history generation unit 122 stores the access result data [acquired] into the access result data [held] (S 36 ), and returns to S 31 .
- the update history generation unit 122 records a result of comparison between the access result data [acquired] and the access result data [held], into the update history (S 35 ). Thereafter, the update history generation unit 122 stores the access result data [acquired] as the access result data [held] (S 37 ). That is, the update history generation unit 122 stores the access result data [acquired] into a predetermined area of the storage unit 13 as data for comparison with access result data to be acquired next. Thereafter, the update history generation unit 122 proceeds to S 38 .
- the update history generation unit 122 ends the processing, and if the number of times comparison has been made between sets of access result data has not reached the upper limit (No at S 38 ), the update history generation unit 122 returns to S 31 .
- the update history generation unit 122 generates an update history for each seed URL as described above.
- a classifier that classifies websites having a possibility of being abused in the future is generated by use of feature values that are update histories of access result data for websites, the update histories including states of management by administrators of the websites. As a result, the classification apparatus 10 is able to accurately perform classification of whether or not each website is a website having a possibility of being abused in the future.
- the generation of a classifier and the classification of websites by use of the classifier are performed by the classification apparatus 10 , but they are not necessarily performed by the classification apparatus 10 .
- the classification apparatus 10 may perform only the generation of a classifier, and the classification of websites by use of the classifier may be performed by another apparatus.
- the information processing apparatus is able to be caused to function as the classification apparatus 10 .
- the information processing apparatus referred to herein includes a desktop or laptop personal computer.
- the information processing apparatus may instead be: a mobile communication terminal, such as a smartphone, a cellular phone, or a personal handyphone system (PHS); or a personal digital assistant (PDA).
- the functions of the classification apparatus 10 may be implemented by a cloud server.
- a computer 1000 has, for example, a memory 1010 , a CPU 1020 , a hard disk drive interface 1030 , a disk drive interface 1040 , a serial port interface 1050 , a video adapter 1060 , and a network interface 1070 . These units are connected to one another via a bus 1080 .
- the memory 1010 includes a read only memory (ROM) 1011 and a random access memory (RAM) 1012 .
- the ROM 1011 stores therein, for example, a boot program, such as a basic input output system (BIOS).
- BIOS basic input output system
- the hard disk drive interface 1030 is connected to a hard disk drive 1090 .
- the disk drive interface 1040 is connected to a disk drive 1100 .
- An attachable and detachable storage medium, such as a magnetic disk or an optical disk, for example, is inserted into the disk drive 1100 .
- a mouse 1110 and a keyboard 1120 are connected to the serial port interface 1050 .
- a display 1130 for example, is connected to the video adapter 1060 .
- the hard disk drive 1090 stores therein, for example, an OS 1091 , an application program 1092 , a program module 1093 , and program data 1094 .
- the various data and information according to the embodiment described above are, for example, stored in the hard disk drive 1090 , and/or the memory 1010 .
- the CPU 1020 loads the program module 1093 and/or the program data 1094 stored in the hard disk drive 1090 , into the RAM 1012 as necessary, and executes the above described procedures.
- the program module 1093 and the program data 1094 related to the classification program are not necessarily stored in the hard disk drive 1090 , and for example, may be stored in an attachable and detachable storage medium and read out by the central processing unit (CPU) 1020 via the disk drive 1100 or the like. Or, the program module 1093 and the program data 1094 related to the program may be stored in another computer connected via a network, such as a local area network (LAN) or a wide area network (WAN), and read out by the CPU 1020 via the network interface 1070 .
- LAN local area network
- WAN wide area network
Abstract
Description
- Non-Patent Literature 1: D. Canali, et al., “Prophiler: A Fast Filter for the Large-Scale Detection of Malicious Web Pages”, in World Wide Web Conference (WWW), 2011.
- Non-Patent Literature 2: C. Ludl, et al., “On the Effectiveness of Techniques to Detect Phishing Sites”, in Conference on Detection of Intrusions and Malware and Vulnerability Assessment (DIMVA), 2007.
- Non-Patent Literature 3: K. Borgolte, et al., “Meerkat: Detecting Website Defacements through Image-based Object Recognition”, in USENIX Security Symposium, 2015.
- Non-Patent Literature 4: K. Soska, et al., “Automatically Detecting Vulnerable Websites Before They Turn Malicious”, in USENIX Security Symposium, 2014.
- Document 1: Internet Archive, “Wayback Machine”, https://archive.org/web/
- Document 2: OpenDNS, “PhishTank”, https://www.phishtank.com/Document
- 3: Malwarebytes, “hpHosts”, https://www.hosts-file.net/
- Document 4: Wappalyzer, https://wappalyzer.com/
-
- 10 CLASSIFICATION APPARATUS
- 11 INPUT AND OUTPUT UNIT
- 12 CONTROL UNIT
- 13 STORAGE UNIT
- 121 CRAWLER UNIT
- 122 UPDATE HISTORY GENERATION UNIT
- 123 CLASSIFIER GENERATION UNIT
- 124 CLASSIFICATION UNIT
Claims (7)
Applications Claiming Priority (4)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
JP2017243914 | 2017-12-20 | ||
JPJP2017-243914 | 2017-12-20 | ||
JP2017-243914 | 2017-12-20 | ||
PCT/JP2018/035987 WO2019123757A1 (en) | 2017-12-20 | 2018-09-27 | Classification device, classification method, and classification program |
Publications (2)
Publication Number | Publication Date |
---|---|
US20200372085A1 US20200372085A1 (en) | 2020-11-26 |
US11423099B2 true US11423099B2 (en) | 2022-08-23 |
Family
ID=66992565
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
US16/955,083 Active 2038-11-12 US11423099B2 (en) | 2017-12-20 | 2018-09-27 | Classification apparatus, classification method, and classification program |
Country Status (4)
Country | Link |
---|---|
US (1) | US11423099B2 (en) |
EP (1) | EP3705974B1 (en) |
JP (1) | JP6823201B2 (en) |
WO (1) | WO2019123757A1 (en) |
Families Citing this family (3)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
US10936677B2 (en) * | 2018-11-28 | 2021-03-02 | Paypal, Inc. | System and method for efficient multi stage statistical website indexing |
US11799876B2 (en) * | 2019-10-04 | 2023-10-24 | Zscaler, Inc. | Web crawler systems and methods to efficiently detect malicious sites |
TW202240426A (en) * | 2021-04-14 | 2022-10-16 | 阿物科技股份有限公司 | Method and system for behavior vectorization of information de-identification |
Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005208814A (en) | 2004-01-21 | 2005-08-04 | Nec Corp | Falsification detection controller, falsification detection control system, falsification detection control method, program and recording medium |
US20110145435A1 (en) * | 2009-12-14 | 2011-06-16 | Microsoft Corporation | Reputation Based Redirection Service |
US20140059203A1 (en) * | 2012-08-23 | 2014-02-27 | Sap Ag | Prevention of coalition attacks in social network communities |
US20150200962A1 (en) | 2012-06-04 | 2015-07-16 | The Board Of Regents Of The University Of Texas System | Method and system for resilient and adaptive detection of malicious websites |
US20160364567A1 (en) * | 2014-03-13 | 2016-12-15 | Nippon Telegaph And Telephone Corporation | Monitoring device, monitoring method and monitoring program |
US9621576B1 (en) * | 2014-12-31 | 2017-04-11 | EMC IP Holding Company LLC | Detecting malicious websites |
US20170293477A1 (en) * | 2014-10-14 | 2017-10-12 | Nippon Telegraph And Telephone Corporation | Analysis device, analysis method, and analysis program |
US20170359368A1 (en) * | 2016-06-09 | 2017-12-14 | Rapid7, Inc. | Identifying malicious identifiers |
US20180041530A1 (en) * | 2015-04-30 | 2018-02-08 | Iyuntian Co., Ltd. | Method and system for detecting malicious web addresses |
US20180048673A1 (en) * | 2016-08-15 | 2018-02-15 | RiskIQ, Inc. | Techniques for determining threat intelligence for network infrastructure analysis |
US20180309772A1 (en) * | 2015-10-19 | 2018-10-25 | Korea Institute Of Science And Technology Information | Method and device for automatically verifying security event |
US20190356675A1 (en) * | 2017-01-11 | 2019-11-21 | Nippon Telegraph And Telephone Corporation | Combining apparatus, combining method, and combining program |
-
2018
- 2018-09-27 US US16/955,083 patent/US11423099B2/en active Active
- 2018-09-27 WO PCT/JP2018/035987 patent/WO2019123757A1/en unknown
- 2018-09-27 EP EP18891675.3A patent/EP3705974B1/en active Active
- 2018-09-27 JP JP2019560805A patent/JP6823201B2/en active Active
Patent Citations (12)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
JP2005208814A (en) | 2004-01-21 | 2005-08-04 | Nec Corp | Falsification detection controller, falsification detection control system, falsification detection control method, program and recording medium |
US20110145435A1 (en) * | 2009-12-14 | 2011-06-16 | Microsoft Corporation | Reputation Based Redirection Service |
US20150200962A1 (en) | 2012-06-04 | 2015-07-16 | The Board Of Regents Of The University Of Texas System | Method and system for resilient and adaptive detection of malicious websites |
US20140059203A1 (en) * | 2012-08-23 | 2014-02-27 | Sap Ag | Prevention of coalition attacks in social network communities |
US20160364567A1 (en) * | 2014-03-13 | 2016-12-15 | Nippon Telegaph And Telephone Corporation | Monitoring device, monitoring method and monitoring program |
US20170293477A1 (en) * | 2014-10-14 | 2017-10-12 | Nippon Telegraph And Telephone Corporation | Analysis device, analysis method, and analysis program |
US9621576B1 (en) * | 2014-12-31 | 2017-04-11 | EMC IP Holding Company LLC | Detecting malicious websites |
US20180041530A1 (en) * | 2015-04-30 | 2018-02-08 | Iyuntian Co., Ltd. | Method and system for detecting malicious web addresses |
US20180309772A1 (en) * | 2015-10-19 | 2018-10-25 | Korea Institute Of Science And Technology Information | Method and device for automatically verifying security event |
US20170359368A1 (en) * | 2016-06-09 | 2017-12-14 | Rapid7, Inc. | Identifying malicious identifiers |
US20180048673A1 (en) * | 2016-08-15 | 2018-02-15 | RiskIQ, Inc. | Techniques for determining threat intelligence for network infrastructure analysis |
US20190356675A1 (en) * | 2017-01-11 | 2019-11-21 | Nippon Telegraph And Telephone Corporation | Combining apparatus, combining method, and combining program |
Non-Patent Citations (7)
Title |
---|
Borgolte. K., et al., "Meerkat: Detecting Website Defacements through Image-based Object Recognition," The Proceedings of the 24th USENIX Security Symposium, Aug. 12-14, 2015, pp. 595-610. |
Canali. D., et al., "Prophiler: A Fast Filter for the Large-Scale Detection of Malicious Web Pages," International World Wide Web Conference Committee, Mar. 28-Apr. 1, 2011, 10 pages. |
Chiba et al, "DomainProfiler: Discovering Domain Names Abused in Future", 46th Annual IEEE/IFIP International Conference on Dependable System and Networks, 2016. * |
Chiba. D., et al., "DomainProfiler: Discovering Domain Names Abused in Future," 46th Annual IEEE/IFIP International Conference on Dependable Systems and Networks, 2016, pp. 491-502. |
International Search Report and Written Opinion dated Nov. 27, 2018 for PCT/JP2018/035987 filed on Sep. 27, 2018, 10 pages including English Translation of the International Search Report. |
Ludl. C., et al., "On the Effectiveness of Techniques to Detect Phishing Sites," International Conference on Detection of Intrusions and Malware, and Vulnerability Assessment, 2007, 20 pages. |
Soska K., and Christin N., "Automatically Detecting Vulnerable Websites Before They Turn Malicious," The Proceedings of the 23rd USENIX Security Symposium, Aug. 20-22, 2014, pp. 625-640. |
Also Published As
Publication number | Publication date |
---|---|
JPWO2019123757A1 (en) | 2020-04-02 |
EP3705974A1 (en) | 2020-09-09 |
EP3705974A4 (en) | 2021-09-01 |
WO2019123757A1 (en) | 2019-06-27 |
JP6823201B2 (en) | 2021-01-27 |
EP3705974B1 (en) | 2022-12-07 |
US20200372085A1 (en) | 2020-11-26 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
US10972493B2 (en) | Automatically grouping malware based on artifacts | |
US11277423B2 (en) | Anomaly-based malicious-behavior detection | |
US9998484B1 (en) | Classifying potentially malicious and benign software modules through similarity analysis | |
US10200390B2 (en) | Automatically determining whether malware samples are similar | |
US9300682B2 (en) | Composite analysis of executable content across enterprise network | |
CN114679329B (en) | System for automatically grouping malware based on artifacts | |
Jiang et al. | A deep learning based online malicious URL and DNS detection scheme | |
US10122742B1 (en) | Classifying software modules based on comparisons using a neighborhood distance metric | |
CN107066883A (en) | System and method for blocking script to perform | |
US10601847B2 (en) | Detecting user behavior activities of interest in a network | |
US11423099B2 (en) | Classification apparatus, classification method, and classification program | |
EP3531329B1 (en) | Anomaly-based-malicious-behavior detection | |
EA038063B1 (en) | Intelligent control system for cyberthreats | |
EP3486809A1 (en) | Classification device, classification method, and classification program | |
Villalba et al. | Ransomware automatic data acquisition tool | |
Alam et al. | Looking beyond IoCs: Automatically extracting attack patterns from external CTI | |
Bhagwat et al. | Detection of ransomware attack: A review | |
CN115643044A (en) | Data processing method, device, server and storage medium | |
JP2016525750A (en) | Identifying misuse of legal objects | |
JP7140268B2 (en) | WARNING DEVICE, CONTROL METHOD AND PROGRAM | |
EP3964987A1 (en) | Learning device, determination device, learning method, determination method, learning program, and determination program | |
Bo et al. | Tom: A threat operating model for early warning of cyber security threats | |
WO2022219792A1 (en) | Collection device, collection method, and collection program | |
US20180373870A1 (en) | System and methods thereof for logical identification of malicious threats across a plurality of end-point devices (epd) communicatively connected by a network | |
KR20230072750A (en) | Method and apparatus for preventing forgery and tampering of sites and files by web threats |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
FEPP | Fee payment procedure |
Free format text: ENTITY STATUS SET TO UNDISCOUNTED (ORIGINAL EVENT CODE: BIG.); ENTITY STATUS OF PATENT OWNER: LARGE ENTITY |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
AS | Assignment |
Owner name: NIPPON TELEGRAPH AND TELEPHONE CORPORATION, JAPAN Free format text: ASSIGNMENT OF ASSIGNORS INTEREST;ASSIGNORS:TAKATA, YUTA;AKIYAMA, MITSUAKI;YAGI, TAKESHI;REEL/FRAME:054872/0324 Effective date: 20201002 |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NON FINAL ACTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: RESPONSE TO NON-FINAL OFFICE ACTION ENTERED AND FORWARDED TO EXAMINER |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: FINAL REJECTION MAILED |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: DOCKETED NEW CASE - READY FOR EXAMINATION |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: NOTICE OF ALLOWANCE MAILED -- APPLICATION RECEIVED IN OFFICE OF PUBLICATIONS |
|
STPP | Information on status: patent application and granting procedure in general |
Free format text: PUBLICATIONS -- ISSUE FEE PAYMENT VERIFIED |
|
STCF | Information on status: patent grant |
Free format text: PATENTED CASE |